Edge Tts

基于微软 Edge TTS 免费实现多语种文本转语音，支持多中文及英文神经声音，无需 API Key。

Security Scan

The skill is primarily documentation + a small install script that only sets OpenClaw configuration to use an 'edge' TTS provider; its requested footprint is consistent with its stated purpose.

Purpose & Capability

Name/description claim a free Microsoft Edge-based TTS and the package only contains docs and a script that configures OpenClaw to use an 'edge' provider — this is consistent if the OpenClaw runtime already implements the provider. The skill does not request unrelated capabilities or secrets.

Instruction Scope

SKILL.md and README only instruct cloning the repo (placeholder URL), setting OpenClaw config keys, and restarting the gateway. No instructions ask the agent to read unrelated files, exfiltrate data, or contact unexpected endpoints. The docs link to external resources (Edge TTS library and Microsoft voice gallery) which is appropriate for the purpose.

Install Mechanism

There is no automated install spec — an included install.sh simply checks for the openclaw CLI and runs openclaw config/gateway commands. Nothing is downloaded from unknown URLs or written beyond standard OpenClaw config changes.

Credentials

The skill requests no environment variables, no credentials, and no config paths beyond editing the user's OpenClaw config; this is proportionate to a configuration-only TTS helper.

Persistence & Privilege

The skill is not marked always:true and does not modify other skills or system-wide settings beyond OpenClaw's own TTS config keys. It does set messages.tts.auto to 'always' by default (user-visible behavior) but that is a configuration choice rather than elevated privilege.

Assessment

This skill is mostly documentation and a small script that flips OpenClaw config to use an 'edge' TTS provider — it does not request credentials or install third-party code. Before installing: (1) verify the repo URL you clone (the SKILL.md uses a placeholder), (2) confirm your OpenClaw runtime actually implements an 'edge' provider (otherwise these config changes won't enable functionality), (3) be aware enabling messages.tts.auto = "always" will cause all replies to be converted to audio (privacy/notification impact), and (4) note the skill states it uses Microsoft services — network calls to Microsoft endpoints will occur when TTS runs and may be subject to rate limits or terms of service. If you want to be extra cautious, inspect the repository you clone and don't run the install script from unknown sources.

Edge TTS Skill - 免费语音合成

🎤 使用微软 Edge TTS 实现免费的文本转语音功能，无需 API Key！

---

📦 安装

# 本地安装
cd ~/.openclaw/workspace/skills
git clone <repo-url> edge-tts

# 或通过 clawhub (如果已发布)
openclaw skills install edge-tts

---

⚙️ 配置

方式一：通过 CLI 配置

# 启用 Edge TTS
openclaw config set messages.tts.provider edge
openclaw config set messages.tts.auto always
openclaw config set messages.tts.providers.edge.enabled true
openclaw config set messages.tts.providers.edge.voice zh-CN-XiaoxiaoNeural

# 重启 Gateway
openclaw gateway restart

方式二：手动编辑配置

编辑 ~/.openclaw/openclaw.json，添加/修改以下配置：

{
  "messages": {
    "tts": {
      "auto": "always",
      "provider": "edge",
      "providers": {
        "edge": {
          "enabled": true,
          "voice": "zh-CN-XiaoxiaoNeural"
        }
      }
    }
  }
}

然后重启 Gateway：

openclaw gateway restart

---

🎤 可选语音

中文语音

语音 ID	描述	性别
zh-CN-XiaoxiaoNeural	温柔女声	女
zh-CN-YunxiNeural	阳光男声	男
zh-CN-YunjianNeural	成熟男声	男
zh-CN-XiaoyiNeural	活泼女声	女
zh-CN-XiaochenNeural	知性女声	女
zh-CN-liaoning-XiaobeiNeural	东北话	女
zh-CN-shaanxi-XiaoniNeural	陕西话	女

英文语音

语音 ID	描述
en-US-JennyNeural	美国女声
en-US-GuyNeural	美国男声
en-GB-SoniaNeural	英国女声
en-GB-RyanNeural	英国男声

其他语言

完整语音列表参考：https://speech.microsoft.com/portal/voicegallery

---

🧪 测试

# 使用 tts 工具测试
openclaw tts "你好，我是 Edge TTS"

# 或在聊天中直接说话，会自动触发 TTS (如果 auto: always)

---

📊 配置说明

配置项	说明	默认值
messages.tts.auto	TTS 触发模式	always / never / mention
messages.tts.provider	默认 TTS 供应商	edge
messages.tts.providers.edge.enabled	是否启用 Edge TTS	true
messages.tts.providers.edge.voice	语音 ID	zh-CN-XiaoxiaoNeural

---

💡 使用场景

• ✅ 个人助手 - 让 AI 助手用语音回复
• ✅ 无障碍辅助 - 为视障用户提供语音输出
• ✅ 多语言支持 - 支持 100+ 种语言
• ✅ 成本敏感 - 完全免费，无 API 调用限制

---

⚠️ 注意事项

1. 网络连接 - Edge TTS 需要访问微软服务
2. 速率限制 - 微软可能有隐式速率限制，大量使用需注意
3. 音频格式 - 输出为音频文件，通过消息平台发送
4. 自动播放 - 部分平台/客户端可能需要用户手动播放音频

---

🔧 故障排查

TTS 不工作

# 1. 检查配置
openclaw config get messages.tts

# 2. 检查 Gateway 状态
openclaw gateway status

# 3. 查看日志
openclaw gateway logs --follow

# 4. 运行诊断
openclaw doctor --fix

常见错误

错误	原因	解决方案
microsoft: no provider registered	Edge TTS 未正确配置	确保 messages.tts.providers.edge.enabled 为 true
音频无法播放	平台不支持	检查消息平台是否支持音频附件
语音不对	语音 ID 错误	确认语音 ID 拼写正确

---

📚 相关资源

• OpenClaw TTS 文档
• 微软 Edge TTS 语音列表
• Edge TTS Python 库

---

📝 更新日志

• 2026-04-07 - 初始版本，支持 Edge TTS 配置和中文语音

---

📄 许可证

MIT License - 详见 LICENSE 文件

作者：Roxy (洛琪希) 🐾